Lithium bank failure and repair possibilities

[markcouz]

I was recently asked to take a look at a neighbours boat where he is having issues with his electric propulsion system - 10kw motor driven by 72v (24s) 280Ah lifepo4 bank.

On 1st inspection I found that the Daly 100A BMS was reporting cells 1>4 at 5v and cells 5>24 at +- 3.28v. I had trouble believing this and on testing I found cells 1>4 to be effectively 0V! (low mV) and cells 5>24 correctly reported at +- 3.28v with very little variance.

He had the 19v controller taking power off 0v and Cell 6 giving approximately 19v which would have caused an imbalance over time that the BMS may have had problems dealing with, however if this had been the cause of the failure I would expect cells 1>6 to have been compromised, not just 1>4 (5 and 6 are testing similar to the rest).

The above issue will be addressed as part of the repair process so that no power is being taken off mid-bank.

After some head scratching I have arrived at the theory that some form of BMS failure has resulted in the BMS over reading the voltage of cells 1>4 (5v), and attempting to balance this by drawing power off of these cells through passive balancing resulting in 0v with 5v still reported.

While importing a replacement BMS via Amazon is possible, I suspect that importing 4x 280Ah cells may be very difficult and hellish expensive!

So here are a couple of questions where I would appreciate some input;

1. Is my mode of failure theory reasonable, or do you see other considerations?

2. My plan is to attempt to recharge Cells 1>2 and then cells 3>4 using a DSP5020 power supply set to 6.56v:5A using the lithium house bank as source, but I have never attempted recharging an LiFePO4 cell from 0v - it this even feasible and are there any associated dangers (bearing in mind that charge power is limited to 33w)?

3. In the unlikely event that recharge is accomplished, how likely is it that these cells may pose a safety risk going forward?

[Statistical]

I would not try to reuse the cells but your plan to charge them with a bench power supply is fine. The problem is limited to 5A it is going to take 280/5= 56 hours. Personally I would the two in parallel and charge it that way to top balance them. Once they are reasonably charged and showing similar voltage I would put all four cells in parallel to final balancing.

You likely don't want to hear this but I would put the Daly BMS in the trash. It has a well deserved reputation for being garbage. Also a 100A BMS (which for Daly is realistically 80A) for electric propulsion requiring 130A is dubious.

Seems crazy your friend spent some serious coin on electric propulsion, 24 high capacity cells, various switches and other wiring costs and then capped it off with the crappiest BMS available.

[markcouz]

Originally Posted by Statistical

I would not try to reuse the cells but your plan to charge them with a bench power supply is fine. The problem is limited to 5A it is going to take 280/5= 56 hours. Personally I would the two in parallel and charge it that way to top balance them. Once they are reasonably charged and showing similar voltage I would put all four cells in parallel to final balancing.

You likely don't want to hear this but I would put the Daly BMS in the trash. It has a well deserved reputation for being garbage. Also a 100A BMS (which for Daly is realistically 80A) for electric propulsion requiring 130A is dubious.

Seems crazy your friend spent some serious coin on electric propulsion, 24 high capacity cells, various switches and other wiring costs and then capped it off with the crappiest BMS available.

Yes the BMS choice was dodgy but apparently supplied as a kit from Canadian source. I made an error quoting 100amp BMS, it is a 150amp. I have recommended going for a 200amp BMS for the replacement because I hate running suspect electronics at anywhere near the rating.

The reason for the slow charge is because his house bank and limited solar array must be able to support the load - the total time will be 56hr x2 because I will be charging 2 at a time (which I could do at 20 amps but will limit to 5 amps because the one thing I have no shortage of is time!)

[tanglewood]

Originally Posted by markcouz

I was recently asked to take a look at a neighbours boat where he is having issues with his electric propulsion system - 10kw motor driven by 72v (24s) 280Ah lifepo4 bank.

On 1st inspection I found that the Daly 100A BMS was reporting cells 1>4 at 5v and cells 5>24 at +- 3.28v. I had trouble believing this and on testing I found cells 1>4 to be effectively 0V! (low mV) and cells 5>24 correctly reported at +- 3.28v with very little variance.

He had the 19v controller taking power off 0v and Cell 6 giving approximately 19v which would have caused an imbalance over time that the BMS may have had problems dealing with, however if this had been the cause of the failure I would expect cells 1>6 to have been compromised, not just 1>4 (5 and 6 are testing similar to the rest).

The above issue will be addressed as part of the repair process so that no power is being taken off mid-bank.

After some head scratching I have arrived at the theory that some form of BMS failure has resulted in the BMS over reading the voltage of cells 1>4 (5v), and attempting to balance this by drawing power off of these cells through passive balancing resulting in 0v with 5v still reported.

While importing a replacement BMS via Amazon is possible, I suspect that importing 4x 280Ah cells may be very difficult and hellish expensive!

So here are a couple of questions where I would appreciate some input;

1. Is my mode of failure theory reasonable, or do you see other considerations?

2. My plan is to attempt to recharge Cells 1>2 and then cells 3>4 using a DSP5020 power supply set to 6.56v:5A using the lithium house bank as source, but I have never attempted recharging an LiFePO4 cell from 0v - it this even feasible and are there any associated dangers (bearing in mind that charge power is limited to 33w)?

3. In the unlikely event that recharge is accomplished, how likely is it that these cells may pose a safety risk going forward?

Maybe it's the font doing strange things, but I don't understand your nomenclature. For example, "Cells 1>2". I don't understand what that means.

[HeywoodJ]

For the systems we work with (Lishen, BYD, CATL) all of the vendors have a recovery cutoff of 1.6V. Above 1.6V attempt to charge and balance; below 1.6V replace the cell. Don't know the validity of that, but that is their protocol/recommendation when doing warranty work.

Originally Posted by tanglewood

Maybe it's the font doing strange things, but I don't understand your nomenclature. For example, "Cells 1>2". I don't understand what that means.

The OP is using the "greater than" symbol as an arrow to say "1 through 4" or "5 through 24" or, in your example 1 through 2. Some piece of the software chain doesn't like it.

 
 1>4 at 5v and cells 5>24 at +- 3.28v
 1>4 at 5v and cells 5>24 at +- 3.28v

The OP is proposing to build a 2S string to hopefully reduce charging time, this should work at least for the bulk part (if charging works at all). At the top end you will need to separate them and balance them, as well as balancing them with their siblings.

Since the 5020 supports an output down to the single cell voltage I would probably parallel two (or even all four) of the affected cells if that is easy to do (perhaps you are trying to not disassemble?) to avoid finding out that some of the cells are recoverable while another in series is not.

[wholybee]

You might find the BMS is fine after you connect it to a good bank without the mid battery tap. Or, you might find a bad connection on one of the leads. Cell readings are sometimes wonky if there is a floating voltage in the middle of the bank, or if there is another path to ground(via the 19V circuit). Not saying it isn't damaged, but worth double checking it.

On this forum and in the marine industry in general, the Daly BMS's are frowned upon. But they are made for traction applications and used for EV conversions and such. It probably isn't as bad of a choice as many would have you believe. A step up in current rating would never be a bad idea.

[markcouz]

Thanks for the comments, next time I post I will avoid the [greater than] symbol!

I have decided based on the comments to charge the 4 cells in parallel at 15amps (I dont like getting too close to max ratings on these types of devices).

If charging is successfull I will check BMS wiring and then reconnect to see if any weird readings are evident before putting it into use or replacing.

[smac999]

I would charge those batteries at 1a. and far away from any house or building, or boat.


though I would also never run a battery bank over 48v either... so apparently your friend likes high risk.